# Arbitrage：利用优势感知投机实现高效推理

- 来源：Apple Machine Learning Research（RSS）
- 发布时间：2026-08-07 08:00
- AIHOT 分数：51
- AIHOT 链接：https://aihot.virxact.com/items/cmsjjjvzz0bczroo5piheqn26
- 原文链接：https://machinelearning.apple.com/research/arbitrage-efficient-reasoning

## AI 摘要

现代大语言模型通过长思维链实现强大推理能力，但推理计算成本高昂。投机解码（Speculative Decoding）用快速但不精确的草稿模型提议 token，再由更强的目标模型并行验证，以加速推理。然而，语义等价步骤中的 token 不匹配会导致不必要的拒绝，传统 token 级投机解码因此受限。

## 正文

Modern Large Language Models achieve impressive reasoning capabilities with long Chain of Thoughts, but they incur substantial computational cost during inference, and this motivates techniques to improve the performance-cost ratio. Among these techniques, Speculative Decoding accelerates inference by employing a fast but inaccurate draft model to auto-regressively propose tokens, which are then verified in parallel by a more capable target model. However, due to unnecessary rejections caused by token mismatches in semantically equivalent steps, traditional token-level Speculative Decoding…